Buffers and buffering power

This chapter is relevant to Section J1(iii) of the 2023 CICM Primary Syllabus, which expects the exam candidates to "describe the chemistry of buffer mechanisms and explain their roles in the body". From the perspective of exam relevance, this probably has the greatest value to the First Part candidate, as many questions have explored this topic in the past:

• Question 3 from the second paper of 2008
• Question 22 from  the first paper of 2014
• Question 11 from the first paper of 2020

Fortunately, this issue is well-described in other chapters, and links to these are offered below as subheadings:

Definition of buffering

• A buffer  is a solution that can resist pH change upon the addition of an acidic or basic components. Buffer solutions consist of a weak acid and its conjugate base::
  • A (weak acid) ↔ A^- (conjugate base) + H⁺
• "Open buffers" are those systems that violate the law of mass action, i.e. where either side of the equation can have some escape or ingress of reagents. An example is CO₂, which can leave the reaction by means of ventilation.

Buffering capacity depends on:

• Concentration of the buffer
• pKa of the buffer (most effective at a pH close to their pKa, where they are 50% ionised)
• Whether the buffer is "open"

Total buffering

• For the whole system, if it were closed (anephric and apnoeic), the buffer capacity (β) is probably about 10mEq/pH/kg body mass, consisting of:
  • 60% intracellular buffering
    • Intracellular proteins (25%)
    • Intracellular peptides (25%)
    • Intracellulr phosphate (10%)
  • 40% extracellular buffering
    • Haemoglobin (20%)
    • Bicarbonate (10%)
    • Plasma proteins (9.4%)
    • Phosphate (0.6%)
• Of these, the bicarbonate buffer system is quanititaitvely the most important.

Intracellular buffering

• Intracellular proteins
  • Present in high concentrations, and full of imidazole histidine residues
  • The pKa of imidazole  (6.8) closely resembles the pH of the cell, making this the optimal buffer
  • Intracellular proteins are present in the greatest concentration (20-30% of intracellular fluid by weight), making them the most important intracellular buffer system
  • 60-70% of total body fluid buffering is due to intracellular proteins (if you count haemoglobin)
• Intracellular phosphate
  • Contributes to buffering more than it does extracelularly
  • High concentration and correct pKa to be a good buffer system
  • However, labile (tends to bind other molecules) and therefore unreliable
  • Contribution is likely much smaller than raw measurements of phosphate would recommend

Extracellular buffering:

• Haemoglobin (the total buffering power of haemoglobin is thought to account for up to 50-60% of the total buffering capacity of the blood)
  • technically intracellular, but does most of the work of buffering the extracellular fluid because of its enormous buffering power.
  • Haemoglobin has many histidine residue groups (pKa close to 6.8)
  • This allows it to act as a buffer at physiological pH, as these histidine molecules act as proton acceptors
  • This buffering capacity increases in its deoxygenated state.
  • As haemoglobin is one of the most populous proteins, it contributes a large proportion of the total buffering in the extracellular fluid, and this is incorporated into the calculation of the standard base excess.
• Bicarbonate is a strong open-ended buffer
  • Available in high concentration
  • pKa 6.1, not exactly close to the pH of body fluid, but good enough because of the high concentration
  • Independently regulated, i.e. can be eliminated by the kidneys, or converted to CO₂ and eliminated via the lungs
  • Buffering potential is maximised by the Hamburger effect:
    • Sequestering chloride into erythrocytes, and increasing the bicarbonate concentration in the peripheral blood.
    • Mitigates the change in pH which would otherwise occur in the peripheral circulation due to metabolic byproducts (mainly CO₂).
  • May not be the highest concentration, but is the most important physiologically overall, because it is open, with massive capacity
• Plasma proteins
  • weak buffer system
  • only present in small concentration (round 60-70g/L)
  • relatively poor histidine content
  • contribute perhaps 20% of the total buffering
• Plasma phosphate
  • low concentration, but correct pKa (close to 7.0)
  • Unimportant except in buffering the urine
• Bone carbonate
  • Long term buffering of chronic acid base disorders
  • Phosphates and carbonates in bone act a supply of slow release buffers
  • This is dependent on renal function to maintain an open buffer system by allowing the elimination of the liberated calcium and phosphate ions (Lemann et al, 2003)

A classic article by J. McNamara and L.I.G. Worthley (2001) would be the best single reference for this topic, as it is a comprehensive review of buffer systems, and easy to read. Another excellent option would be Chapter 5 from Hasan's Handbook of blood gas/acid–base interpretation (2009), as it is brief and literally intended to be quickly flipped through. One needs to warn the reader that there is no need to burrow deep into the intricacies of acid-base physiology to be able to answer this question to a satisfactory degree. Still, sometimes people come here with a strange need. For those interested  readers whose curiosity is only as strong as their highschool-level chemistry, Ulysses S. Seal's The Chemistry of Buffers (1965) would be satisfactory.

Definition of buffering

"Buffer" seems to be the one term IUPAC have no strict definition for. Borrowing one from  McNamara and Worthley seems reasonable, as the second author has probably written half of the data interpretation SAQs from the first two decades of the CICM exams. In turn, these authors borrowed their definitions from the report of the Ad Hoc
Committee of the New York Academy of Sciences, 1966,  a frequently referred-to piece on acid-base terminology.

"A buffer  is a solution that can resist pH change upon the addition of an acidic or basic components"

It resists this change by being ambivalent about its ionisation. To borrow a turn of phrase from Larewnce Henderson (1908),

 "Other things being equal, the greatest possible efficiency in preserving neutrality, on both sides of the neutral point, is possessed by that acid whose ionization constant is precisely equal to the hydrogen ionization of water divided by the degree of ionization of the salt."

Buffers work best at a pH close to their pKa, so that they are about 50% ionised. In this state the solution is composed of 50% weak acid (A⁻) and the proton H⁺, and 50% its salt (HA), described by this unimaginative relationship:

 HA ⇌ H⁺ + A⁻

For example, consider a perfectly balanced citrate buffer solution, at the pKa of citrate (6.40)  would contain 50% dissolved citric acid and 50% citrate anion, with its associated H⁺. The equilibrium is maintained until  H⁺ is added or removed. The addition of H⁺ increases the number of reagents on the right side of the equation, leading some of the H⁺ to combine with the citrate anion and produce citric acid. If all of the added H⁺ remained in the solution, the solution would drop its pH quite considerably, but some of the H⁺ ends up getting soaked up by the citrate, and the pH fluctuates only slightly, coming to rest at an equilibrium slightly offside from the original value (say, 6.35). This, then, is the basic function of a buffer.

For a more scientific explanation (in case one is confronted with this in some kind of cross-table viva) one could draw a relationship such as this, describing the percentage of ionised components in solution as a function of pH (otherwise known as a Bjerrum plot)

The objective of drawing such a crude graph would be to then point out that within a narrow range of the pKa value, the curve is at its steepest, i.e. the buffer can increase or decrease its ionisation rather dramatically with only a small change in pH. 

Conversely, when the buffer is at a pH which is far from its pKa, it becomes much less capable as a buffer. If only 1% of the buffer is ionised, there are too few molecules to absorb an influx of H⁺, and most of the added H⁺ would remain, producing a sizeable change in pH.  

Unfortunately, it appears there is no avoiding a discussion of Ka and pKa here, and the author engages this subject only with the greatest reluctance, which is not a uncommon reaction, though in this case mainly for reasons related to impostor syndrome.  As the term appears everywhere, with excellent explanations abundant in free and paid literature, one might justifiably outsource the explanation of it to real professionals; but in that case Deranged Physiology would be literally the only website on the internet without an explanation of the acid dissociation constant.  Therefore:

pKa, the negative log₍₁₀₎ of the acid dissociation constant

Predictably, as soon as we mention that the equation (HA ⇌ H⁺ + A⁻ ) is at an equilibrium, we imply that some kind of equilibrium constant must describe it. It would surely have to be a Konstant, because German, described by the relationship

K_a  =[A⁻][H⁺] / [HA]

K_a denotes die Säurekonstante, where the capital K is an equilibrium constant (whereas a lowercase k is by convention a rate constant) and the lowercase subscripted a stands for acid and is not italicised because it does not refer to a number (upright letters only, apparently). To omit the italics and subscript (Ka) is a completely legitimate lazy shortcut, adopted by many publishers in the early stages of the twentieth century because of the difficulty and expense of typesetting subscripted lowercase. These days IUPAC have micromanaged everything, including even this, and the modern era has relatively effortless font control; but it requires more clicks than the author can manage while writing these notes at 3am, and moreover nowhere else in Deranged Physiology is the a under the K. 

Anyway: the Ka value describes the "strength" of an acid, in that it is the dissociation constant of the equilibrium state, and so if the acid is fully dissociated the [HA] will be extremely small and the [A⁻][H⁺] will be extremely large, giving a very large Ka value. The Ka of hydrochloric acid is 1.3 ×10⁶, as an example; and here is a list of other Ka values for some perspective. Obviously at some stage one might start to wonder whether this value can also apply to bases (it does), and whether it would then be  changed to Kb (it would). In fact, 

K_b  =[OH⁻][B⁺] / [BOH] 

where the lower the Kb value, the stronger the base (i.e opposite to the direction of Ka).  In fact there is also Kw, the autoionisation constant of water, which is 1.006×10⁻¹⁴  at 25°C.  

Why, then, do we have pKa and pKb?  The p again stands for "power" or "potenz" and describes the negative logarithm to the base of 10, i.e.

pK_a  = - log₁₀ K_a

But why? It seems the main reason is that there is a great reluctance for chemists to operate unwieldy exponents such as 1.006×10⁻¹⁴ for a property that spans beyond fourteen orders of magnitude. The convention is therefore to operate negative logs. The lower the pKa of an acid, the stronger the acid, and the higher the pKb of a base, the stronger the base. This is also coherent with the scale of pH values, which becomes convenient, because the pKa or pKb of a substance is therefore the pH at which it is exactly 50% ionised. To borrow an excellent mnemonic device from Part One, 

• Bases are ionised Below their pKa
• Acids are ionised Above their pKa

Also, for a given pair of an acid (a) and its conjugate base (b),

pK_a + pK_b = pK_w 

where pKw = 14. In other words, a strong acid must have a weak conjugate base, and vice versa. This is only valid for aqueous solutions, which is fortunate because that's what we happen to be. The reason for this maddening excursion into basic chemistry is because it is fundamental to the understanding of the extent to which a buffer is able to resist the change in pH, or the "buffering capacity" of the buffer.

Buffering capacity (BC) and buffer index (β)

From the above, and in general on the basis of common sense logic, it follows that there must be some relationship between the number of free buffer anions in solution and the capacity of that solution to absorb any added H⁺. More anions means more H⁺ can get absorbed and therefore a greater pH change can be resisted, i.e. the capacity of the solution to act as a buffer is greater. Obviously, there are two main ways one could increase the buffer capacity:

• by being at a pH closest to the pKa of the buffer (so more of the buffer is ionised)
• by having a higher concentration of the buffer substance in solution

This is the basis of the concept of buffer capacity, which is occasionally represented as buffer index (β) since this is the Greek letter used by Donald Van Slyke when he first wrote about this concept in 1922. The definitions used in that 100-year-old paper are almost identical to the modern definitions, but to be absolutely IUPAC-precise, 

Buffer capacity, or buffer index, is "the capacity of a solution to resist changes in pH on the addition of strong acid or strong base which may be expressed numerically as the number of moles of strong acid or strong base required to change the pH by one unit when added to one liter of the specified buffer solution"

In other words, 

β = d A / dpH 

where A is the number of moles of strong acid 

So, where a one litre solution of something requires o.2 mol of concentrated HCl to change from a pH of 7.0 to a pH of 6.0, the buffer index is 0.2/1.0 = 2.0 at the mean pH of 6.5. Buffer capacity and buffer index are actually slightly different as concepts, and though buffer index (β) might be satisfactory for biological solutions with low concentrations of acids and bases, scenarios that call for large pH changes and strong reagents are better suited to using BC, though the author will not embarrass himself by trying to explain in what way or what the significance of this is. The reader with endless appetite for hardcore chemistry is instead redirected to the excellent works such as Chiriac & Balea (1997). For the purposes of the following discussion, composed by and intended for an audience of beings that operate within a very narrow pH range, BC and β can be considered synonymous and will be used interchangeably.

Total body buffering capacity

The reader familiar with the usual operation of Deranged Physiology will have, by this point in the discussion, become impatient with the lack of empirical supporting data, represented by pioneering measurements taken from three dead pigs by a solitary investigator from the 1950s. In this case the most often referred-to paper is by Swan & Pitts, whose 1955 experiment on some unknown number of anephric dogs is sufficiently famous to have been quoted by Kerry Brandis. The investigators infused 10mmol/kg of hydrochloric acid into their animals and determined that, after 160 mmol of HCl, the average 16kg dog dropped their pH from 7.41 to 7.10 and bicarbonate from 24.6 to 7.0 mmol/L. Following the equation above, this gives a buffer capacity of something like 32.2 mEq buffering capacity per every 1kg of dog.

However, the "lightly anaesthetised" animals were left to breathe spontaneously and noticeably hyperventilated, which is cheating. For establishing the true buffer capacity of the body as a closed system, we are grateful to thirteen more anephric dogs, and to Albers et al (1971) who measured their response to different levels of FiCO₂. The d A / dpH was reasonably straightforward to calculate as the dA is known from the dose of CO₂ and the Henderson-Hasselbach equation, and the dpH was measured directly from the blood. For the whole body of the anaesthetised dog, the buffer capacity was approximately 10.7 mEq/kg/pH.

In other words, 10.7 mEq of acid was required to acidify 1kg of dog by 1 unit of pH, or 18 mEq for every 1000ml of body water.  From these data, if we assume that CO₂ converts to an equimolar amount of H⁺, this suggests that a 70kg human with 40L water will go from pH 7.4 to a pH of 6.4 (and nearly dead) with the addition of about 750 mmol of CO₂, which is 16.8 litres of gas at standard temperature and pressure. And as a normal human tends to generate about 10mmol (224 ml) of CO₂ per minute, a 70kg human who is unable to breathe would, on the basis of these data, be expected to drop their pH by about 0.013 mmol/L each minute. In fact this is roughly what is observed in apnoeic anaesthetised patients:

As you can see, each minute (apart from the initial rise)  resulted in a small 0.01-0.02 decrement in pH, which lurks at the limits of the measuring capacity for mass-produced glass electrodes; and the average rate of change was about 0.016/min, which is close to the expected value (allowing some elasticity because none of these patients were anesthetized anephric dogs). 

Albers et al (1971) is an often-quoted paper not only because it is often corroborated by human findings, but also because the authors took special efforts to determine that the contributions to buffering were 4.0 mEq/Kg/pH in the extracellular space and 6.7 mEq/Kg/pH in the intracellular space, when the spaces were measured as wet tissue rather than fluid. The takeaway from this is that the intracellular compartment is capable of buffering about twice as much acid as the extracellular systems, at least on paper. Practically speaking, however, the extracellular fluid is the dominant buffer mechanism:

Extracellular buffering

Extracellular fluid, and specifically blood, is an essential source of buffering capacity, for reasons largely related to the fact that it circulates around the lungs and kidneys. Even when it is confined to a beaker and made motionless, it has a huge buffer capacity, an observation made by Hans Friedenthal (1902) who found about seventy times more sodium hydroxide needed to be added to whole blood than to sterile water in order to produce a measurable change in pH. For contrast, the reader is reminded that pure distilled water is conventionally said to have no buffer capacity, though in reality distilled H₂O will absorb some CO₂ from the atmosphere and develop some trivial buffer capacity as the result.

 Friedenthal only had phenolphtalein and colour change to measure the difference in pH, and later authors were better able to measure the buffering capacity of whole blood and plasma (Ellison Straumfjord and Hummel, 1958). For whole blood, at pH of 7.4, the capacity was 38.5 mEq/L/pH, whereas for plasma alone it was 16.1 mEq/L/pH, suggesting the red cells played a major role in the buffering process.  In plasma, the plasma proteins were by far the most active buffer (8.1 mEq/L/pH), followed by the bicarbonate and phosphate. However, bicarbonate remains quantitatively the most important buffer system in vivo because it is an "open" system. 

The concept of an "open" buffer system

The term "open buffer" only seems to exist in physiology textbooks, enthusiast-grade resources such as Part One, and that that paper by McNamara and Worthley from 2001. It does not appear to have a formal chemical definition as such. Still, the accepted meaning for this term is "a system where reagents can be added or subtracted at one or both ends". In short, it is an unbalanced chemical reaction where the experiment allows cheating. The CO₂ ⇌ bicarbonate pair is one such system, as excess CO₂ can be removed by ventilation, and excess bicarbonate can be removed by renal elimination. These mechanisms are so powerful at resisting pH change that they needed to be disabled for accurate measurements of total body buffering capacity by Albers et al (1971)- the animals were paralysed and ventilated to control the movement of CO₂, and their renal arteries were ligated.

Bicarbonate as the main extracellular buffer system

With intact open buffer mechanisms, the bicarbonate system can accommodate vast quantities of acid. Considering that the production of CO₂ as a metabolic byproduct is a constant 10mmol/min, and that the body fluid CO₂ concentration remains stable, and that CO₂ in water is expected to produce an equimolar amount of H⁺, the bicarbonate system can be said to "neutralise" about 15-20 mol of H⁺ per day, which is quantitatively the largest contribution to buffering among all systems. Considering that this is happening at a normal resting minute volume, and that minute volume could theoretically increase by up to 15-20 times, the capacity of the system to adjust to added acid load must be massive. Yes, the reader may point to the fact that other sources of acid production in the body may also manage large amounts of H⁺, but all of these protons are hidden inside cells and are often consumed immediately in the reverse version of whatever reaction produced them, making their existence a purely theoretical construct for the purpose of discussing acid-base balance. The production of CO₂, on the other hand, is a genuine addition of "exogenous" acid into the equation.

In summary, the bicarbonate buffer system is quantitatively the most important. And, to clumsily tie this section with the next, this system is enhanced by the CO₂ carriage function of haemoglobin, allowing for an even greater buffering power.

Haemoglobin as an extracellular buffer

Haemoglobin acts as a buffer in two main ways:

• By binding CO₂ and removing it from the circulation
• By binding H⁺ i.e. acting as a buffer directly

It's hard to know where to put this material, as the reader revising for an exam may look for it in the section dealing with the Haldane Effect or the Bohr Effect, or where the practicalities of buffering for respiratory acid-base disorders are discussed, or in the section dealing with the structure and function of haemoglobin. Rather than try to centralise the information, it has been scattered in several different directions, and appears in each section in a different way, according to whatever felt most appropriate to the younger version of the author.  That guy had a lot going on, and to forgive him his loose thinking would be the only compassionate response. 

Haemoglobin buffering as part of the Haldane Effect: By binding CO₂ in its deoxygenated state to form carbamate compounds, haemoglobin removes CO₂  from the solution and decreasing the total dissolved CO₂ content of the bloodstream. Incidentally this also has the reciprocal  function, where the oxygen affinity is decreased by CO₂ binding and low pH, allowing the release of carried oxygen into the tissues (which is referred to as the Bohr Effect).  The mechanism of carbamate is discussed in more detail in the chapter on the transport of CO₂ in the blood, as well as in a deep dive by Austen Riggs (1988), but, just briefly,

• CO₂ binds to uncharged N-terminal α-amino groups on both α and β subunits of haemoglobin
• Oxygenation of the haem iron atom in a haemoglobin molecule is a heterotropic allosteric modulator of these CO₂ binding sites because it introduces a conformational change to the haemoglobin tetramer (positive cooperativity)
• As the result of this allosteric modulation, CO₂ has a higher affinity for the deoxygenated T state than for the R-state
• This mechanism contributes 70% of the total CO₂ carriage due to the Haldane effect (Roughton, 1964), and therefore about 10-15% to the total transport of CO₂ in the blood.

Haemoglobin buffering by histidine residues:  By using its plentiful histidine residues, haemoglobin can mop up the H⁺ liberated in the process of CO₂ dissociating into HCO₃^- and H⁺, which it does best when it is in its deoxygenated state. 

• Each haemoglobin tetramer molecule has 38 charged histidine residues, of which four are attached to the haem group.
• The dissociation constant for each of these histidine residues is influenced by the oxygenation of the haem, which is explained with unnecessary diagrams here in the page on the Bohr effect. To cut a long story short, the pKa of the haemoglobin molecule changes due to oxygen binding: the overall pKa of deoxyhemoglobin is said to be 7.71  and the pKa of oxyhemoglobin is around 7.16.
• As a result, when haem loses oxygen, the haemoglobin tetramer as a whole becomes more basic. 
• This removes hydrogen ions from solution (i.e. buffers the solution).
• The effect of this on the equilibrium of bicarbonate and carbonic acid favours the conversion of carbonic acid into bicarbonate.
• Thus, the loss of oxygen from the haemoglobin tetramer, by buffering, increases the amount of CO₂ being carried in the form of bicarbonate.
• This probably contributes only about 30% of the Haldane effect, i.e. the increased capacity of the blood to carry CO₂ under conditions of decreased haemoglobin saturation is mainly the result of carbamate formation and only partly due to the fact that haemoglobin buffering shifts the balance of the Henderson-Hasselbach equilibrium in favour of the reaction products.

The buffering capacity of haemoglobin  is generally said to contribute about 50-60% to the total buffering capacity of blood, which seems to track back to the Ellison Straumjord and Hullen experiment. They found that, of the total buffer capacity of whole blood (38.5 mEq/L/pH), approximately 58% disappeared when the red cells were removed.   How does haemoglobin have such a massive buffering power? Sure, there's about 330g/L of it in the RBC cytosol, but it is a relatively large molecule and therefore present in a fairly small molar concentration (often misrepresented as something like 8.7-11.2 mmol/L, which is actually the total molar concentration of each monomer, whereas the concentration of the tetramer is more like 2-3 mmol/L). Fortunately, the total buffering effect is increased by the advantageous pKa of the multitudes of buffer-capable histidine residues which it carries, meaning that each molecule can absorb more than just an equimolar amount of H⁺. 

And at this stage, one must acknowledge that the chapter on buffering has gone on for quite some pages now without addressing the question of what exactly these histidine residues are, and this seems like something of fundamental importance if the discussion is going to move in the direction of discussing the buffering properties of proteins. Therefore:

The buffering effects of histidine residues

Proteins can act as buffers in solution because they are generally weak acids, i.e. ionised into a mostly anionic protein molecule and some number of H⁺.  Well, they are actually large molecules with a rather unevenly distributed charge on their surface, some of which is negative because of the side chains of amino acids and various other exterior molecular groups. The medical reader is occasionally confused by the term "residues" that usually gets applied to these structures, which is a normal reaction, as it is never clear what this means until you go looking. A "residue" is the semiformal colloquialism to describe what's left of an amino acid molecule following the reaction that combines them together, as this reaction typically results in the loss of some atoms, usually hydrogens and oxygen.  it would therefore be incorrect to continue calling these "amino acids", as the molecule that remains is now different; ergo "residue", the leftovers.

Thus, theoretically the negatively charged "leftovers" of  amino acids can bind H⁺. (though not all residues are ionised, and not all ionised residues are negatively charged). The obsessed reader can be directed to the extensive contributions made by Grimsley et al (2009), whose team compiled the measured pK values for 541 ionisable groups from 78 proteins. A brief glimpse, from Madies & Cohen (1982), lists some of the representative residues and their specific ionisable functional groups: 

Dissociable group	pKa value
α-Carboxyl	3.6-3.8
β-Carboxyl (aspartic acid)	≈4.0
γ-Carboxyl (glutamic acid)	≈4.0
Imidazole (histidine)	6.4-7.0
α-Amino (C and N terminals)	7.4-7.9
Sulfhydryl (cysteine)	≈9.0
-Amino (lysine)	9.8-10.6
Phenolic (tyrosine)	8.5-10.9
Guanidino (arginine)	11.9-13.3

It will suffice to summarise that most of these have a pKa which is well outside the normal physiological range, and that the only amino acid residues that have a useful pKa are histidine residues. And by "useful", we mean a pKa sufficiently close to the pH of the human body fluids, such that significant proportions of the residue exist in both a protonated and unprotonated form.  Specifically, it is the imidazole group of the histidine residues: a small 5-membered ring molecule with a labile relationship with its H⁺: 

The relevant equilibrium is between the positively charged imidazolium ion and neutral imidazole:

The donation of a proton changes the ring charge from +1 to 0. The neutral form still has one N-H bond; its other ring nitrogen can accept a proton. Thus, the protonated form is the proton donor, i.e. the conjugate acid, and the neutral form is its conjugate base (Pogostin et al, 2019).

The pH of 6.0 in the diagram above seems like a nice round value, and from this, the reader should have immediately caught the whiff of oversimplification. In fact the situation is much more complicated, the residues are individually charged , there is no single pKa that applies to every histidine residue, and the surrounding protein structure and local chemical environment can shift this value, changing the proportion of protonated rings at a given pH (Grimsley et al, 2009). Below the relevant pKa, the protonated form predominates; above it, the neutral form predominates; and 6.0 is a conveniently memorable midway value which will seem correct for most reasonable examiners.

If the reader has come here from the page about the alpha-stat method of ABG interpretation, they will connect that this is the structure that defines the "alpha", that being the ratio of protonated to total imidazole residues among the protein molecules (normally about 0.55).  

The aromatic imidazole ring of the histidine molecule is said to have a pKa of 6.8 at normal body temperature. This is an oversimplification which is mostly true, but because the quaternary structure of proteins is subject to so much change, each individual  imidazole ring pKa value will depend on where it is positioned. "Buried" histidines have a markedly different pKa to exposed ones, for example (Edgecomb & Murphy, 2002); moreover proteins can move and shapeshift, burning some histidines and exposing others depending on the circumstances. Several good examples of this exist, most notably haemoglobin, which has a somewhat increased buffering capacity in its deoxygenated state as has already been discussed above. 

Plasma proteins as an extracellular buffer system

Given that most other amino acid side chains are quite useless at buffering anything at physiological pH ranges, the "buffering power" of all proteins is almost entirely related to the number of histidine side chains they have. Thhus, the plasma proteins are rather weak participants in buffering, owing to the fact that they are present in small concentration (only around 60-70g/L) and have low histidine content. It is usually said that they contribute about 20% to the total nonbicarbonate (i.e. protein) buffering capacity of the blood, the rest being supplied by haemoglobin. Textbooks often quote buffer capacity values of up to 1.4 mmol/g/pH for albumin, and up to 0.08 mmol/g/pH for globulin, values which come from Van Slyke (1928) and which amount to something like 17 mEq/L/pH for plasma. More modern measurements by Figge et al (1991) suggest that this figure is probably closer to 12 mEq/L/pH, mostly because globulins don't seem to contribute anything at all. The precise figure is probably unimportant, and the reader is left with the take-home impression that plasma proteins have minimal importance to buffering.

Plasma phosphate as an extracellular buffer

Phosphate (HPO₄^2-) is the anion of phosphoric acid (H₃PO₄), a weak acid with a pKa of 6.8.  At body fluid pH it is usually a mixture of (PO₄^-) and (PO₄^2-), i.e. ionised in one way or another, and therefore capable of buffering H⁺. However, there is very little of it around - approximately 1-2 mmol/L under most normal circumstances. The contribution from this system to extracellular buffering are therefore very limited - according to Ellison Straumfjord and Hummel, it is responsible for 0.6 mEq/L/pH of the total blood buffer capacity, or around 1.5%. The only possible redeeming feature of this system is that it is open at both ends, where phosphate can be entrained back into cells, incorporated into bone, or (most commonly) eliminated by the kidneys, where it buffers the urine and accounts for about 50% of the total really cleared nonvolatile acids. In summary, extracellular phosphate could theoretically act as a solid buffer system, if only it were present in the kind of concentrations that would be enough to kill you. 

On the other hand, inside the cells:

Intracellular buffering

Cells seem like a pretty important component to consider, when it comes to buffering. Originally the whole concept of buffering in biological systems was developed by Lawrence Henderson in the early 20th century on the basis of measurements made in phoshate/bicarbonate solutions designed to simulate the content of cells (Henderson & Black, 1907,  "Concerning the neutrality of protoplasm"). Huge amounts of acid needed to be added to a modest solution of sodium bicarbonate and potassium phosphate with little change in the pH. "The justness of the neutrality of protoplasm must far surpass the accuracy of adjustment of any other known equilibrium in the body," the authors concluded; "accordingly protoplasm is extraordinarily safeguarded .... from variation in hydrogen or hydroxyl ionization". Acknowledging the known challenges of measuring the buffering capacity of living cells (in the sense that one tends to destroy them by impaling them with glass electrodes), a table made of experimental data from Burton (1978) can be presented as a list of representative values:

Of the total buffer capacity, which in man appears to be something like 20-40 mEq/kg/pH in cell water, Burton credits 60% (or about 24 mEq/kg/pH) to protein, and the rest to phosphate, with the caveat that intracellular phosphate can't buffer right now because it's busy.

Intracellular phosphate buffers

The cells contain tons of phosphate. That it is the most abundant of the intracellular anions is an often repeated factoid, which is probably accurate with respect to the total phosphate content of incinerated rats, and probably less so in the living organism, depending on what one means by "abundant". Surely, abundant phosphorus-containing molecules are present in cells, but most of these neither release free phosphate to do buffer work, nor do any buffering themselves. To demonstrate this, Burton (1978), counting all the osmotically active intracellular molecules that could be rounded up in the cytosol, and calculated that if they all buffered in some way then total intracellular buffering capacity should be something like 69 mEq/kg/pH, which is at least three times what is measured experimentally.

This is because free inorganic phosphate of the H₂PO₄^- and HPO₄^2- variety is present in cells only in small and hightly variable concentrations ranging probably from 14 mmol/L to 0.78 mmol/L. Ergo, most of the buffering inside cells is performed by other substances.

Intracellular protein and peptide buffers

The cytosol is full of proteins and peptides, to the tune of 20-30% by mass. Of these a fair number are capable of buffering, again mostly by histidine residues. Castellini & Somero (1981) and  Somero (1985) compiled a list of well-buffered vertebrates and concluded that the total buffering capacity of muscle (where buffering is the most essential) is probably about 50% proteins and 50% dipeptides such as carnosine, anserine and balenine (all of which have histidine as one of the amino acids). The later contributed something like 137 mEq/kg/pH to the total buffer capacity of scombrid fish.  Interestingly, and not unexpectedly, those organisms most likely to enjoy vigorous activity under anaerobic conditions are also the most likely to have the need for massive buffer capacity. Consequently the higher buffering capacities belong to diving mammals such as the spotter porpoise (84 mEq/kg) and fast pelagic fish such as the spotted mackrel (109 mEq/kg). Molluscs, the undisputed kings of anoxic survival, tend to have rather modest muscle buffer capacity, probably because their metabolism is too excellent to produce much acid. 

Bone as a buffer system

Lemann et al (2003), Burton (1991) and Bushinsky & Kriger (2022) discuss bone buffering in much detail, though it is not often mentioned in textbooks and SAQ answers because it does not play much of a role in the course of normal acid-base balancing. In summary, the process of buffering by bone mineral is as follows:

• Bone contains about 80% of the total body bicarbonate stores
• Octocalcium phosphate, a precursor to hydroxyapatite, can dissolve in the presence of a low pH, liberating bicarbonate and calcium
• This can only work as a buffer sustainably if calcium can be eliminated
• Moreover the bone surface is exposed to relatively little ECF flow, which means the process of bone dissolution for buffering is very slow
• Only about about 9L per day is gradually exchanged, and forms 15% of the total daily CO₂ production 

The isohydric principle

The number of buffer mechanisms and buffer-like molecules is enough to confuse even the most determined reader, but it is important to introduce some calm at this stage by reinforcing that only one really needs to ever be measured directly, and this is the CO₂ - bicarbonate system. There are two reasons for this:

1. The CO₂ - bicarbonate system is quantitatively the most important, as it is capable of moving 15,000-20,000 mmol of acid per day, and can scale easily to increase its capacity if the renal and respiratory function remains normal. In contrast, plasma protein and phosphate are neither especially scalable, nor especially capacious, and so measuring them is pointless; and renal elimination of "fixed acids" amounts to perhaps 50-100 mmol/day, in other words it is dwarfed by the magnitude of the CO₂. 
2. It is moreover pointless to measure all the buffer systems individually, as firstly one cannot exactly reach inside the cells to examine the protonation of the imidazole residues, and secondly because they all have one common reagent, which is H⁺. Thus, all dissociated weak acids and bases in the same solution will be at equilibrium, because all of them ultimately adjust themselves to the same H⁺ concentration. This is called the isohydric principle.